NSDann, a LabVIEW Tool for Neutron Spectrometry and Dosimetry Based on the RDANN Methodology

Abstract

Nowadays the ANN theory it is offering an alternative for those problems in which the traditional methods have not given very convincing or not very convenient results. The most successful applications in ANN are: planning, prediction, optimization and recently has been applied in the neutron spectrometry and dosimetry research area. The use of the ANN technology is a useful alternative to solve the neutron spectrometry and dosimetry problems, however, several drawbacks must be solved in order to simplify the use of these procedures. In the present work was developed under the LabVIEW programming enviroment, a computer tool based on AI technology. Starting from the extracted knowledge of an ANN designed by means of the Robust Design of Artificial Neural Networks methodology, which is capable to unfold neutron spectra and to simultaneously calculate 13 equivalent doses. The designed tool presents the following characteristics: was designed in a user’s graphic interface, it is easy to use, is not essential on the part of the user, to have knowledge of ANN neither of neutron spectrometry and dosimetry to operate the tool, the execution speed and performance of the program is high and, contrary to classical existent codes based on iterative algorithms, in the application developed in this work, is not required an initial guess spectrum to obtain the neutron spectra wanted. Finally, this tool needs as entrance data just the count rates coming from the BBS.